Deconvolution of core electron energy loss spectra

Area of Science:

• Materials Science
• Spectroscopy
• Computational Physics

Background:

• Electron energy loss spectroscopy (EELS) is crucial for materials analysis.
• Multiple scattering and instrumental broadening distort EELS spectra, complicating analysis.
• Accurate deconvolution is essential for obtaining reliable single scattering EELS data.

Purpose of the Study:

• To compare the performance of various deconvolution methods for core-loss EELS.
• To identify and analyze artifacts introduced by different deconvolution techniques.
• To evaluate the effectiveness of model-based deconvolution against other methods.

Main Methods:

• Description of Gaussian modifier, Wiener filter, maximum entropy, and model-based deconvolution methods.
• Performance evaluation using simulated EELS spectra with known single scattering distributions.
• Validation of methods on experimental core-loss EELS spectra.

Main Results:

• Model-based deconvolution demonstrated superior performance in accurately recovering the single scattering distribution.
• Comparison revealed significant artifacts associated with Gaussian modifier and Wiener filter methods.
• Maximum entropy methods showed limitations compared to model-based approaches, particularly regarding error estimation.

Conclusions:

• Model-based deconvolution is highly effective for correcting multiple scattering and instrumental broadening in EELS.
• The ability to estimate error bars from single spectrum acquisition is a key advantage of model-based methods.
• This work provides guidance for selecting appropriate deconvolution techniques in EELS analysis.